Color centers as simultaneous active laser media and saturable absorbers

Volume 79, number 1,2 OPTICS COMMUNICATIONS 1 October 1990

Color centers as simultaneous active laser media and saturable absorbers

N.D. Vieira Jr., L.S. Assis and S.P. Morato

Instituto de Pesquisas Energ~ticas a Nucleares, Cornissgto Nacional de Energta Nuclear - SP, Caixa Postal 11040, CEP 05499, S~o Paulo, Brazil

Received 2 January 1990

There is a family of color centers and metal ion doped crystals that absorb in the emission region of the neodimium lasers.

These centers are in general laser active, usually being pumped in a collinear scheme. They can also be used as saturable absorbers for cw passive Q-switching, but this limits the maximum absorption that can be introduced in the laser cavity thus limiting the maximum gain achievable in the saturable absorber/active pumped medium. By using a coupled cavities scheme we were able to optimize the performance of the pumped medium and obtain CW Q-switched operation by an injection mechanism, the saturable absorber being the unique source of loss for the pump. We coupled an astigmatically compensated KCI:TI°( 1 ) color center laser cavity with a Nd laser cavity using both properties of the KCI: T1 ° ( 1 ) crystal. We obtained complete Q switching of both lasers with a typical pulse width of 1 ps and a pumping efficiency of the color center laser in the order of 40% achieving a maximum output power of 1.3 W for an input pump power of 3 W from the Nd: YAG laser. The Nd: YAG Q-switched operation was also tested in this coupled cavities scheme using LiF: F~- color centers. The obtained pulse widths agree well with a simple theoretical model.

I. Introduction

Color center lasers ( C C L ) behave similarly to dye lasers a n d are their c o u n t e r p a r t in the n e a r infrared region [ 1,2]. They are very efficient mostly due to the possibility o f being p u m p e d by a n o t h e r laser in very small areas ( ~ 10 -5 cm2), a n d achieving high gain in very small v o l u m e s ( ~ 10 -6 cm3). I n par- ticular there is a class o f color centers that can be ef- ficiently p u m p e d by N d lasers due to their strong ab- sorption b a n d s in the 1 lain region. Some of these centers are TI°( 1 ) in KC1 [3,4] ( 2 = 1.040 ~tm) a n d F £ in L I F [ 5 , 6 ] ( 2 = 0 . 9 6 ~tm) with b r o a d absorp- tion b a n d s that overlap the N d laser emission. T h e TI °( 1 ) : KC1 center emits in the 1.5 ~tm region a n d m u s t be operated at 77 K. T h e F~- : L i F center emits at 1.12 p m a n d is stable at r o o m t e m p e r a t u r e even u n d e r strong p u m p i n g laser light [ 7 ].

O u r objective was to use these centers as saturable

absorbers a n d gain m e d i a simultaneously. As a sat- urable absorber m e d i u m it controls the temporal be- h a v i o u r o f the N d : YAG laser d e p e n d i n g on its sat- u r a t i o n intensities, a m o u n t of a b s o r p t i o n a n d decay times. As gain m e d i a their decay t i m e d e p e n d s u p o n the i n t r a c a v i t y intensity o f the CCL r a d i a t i o n field.

Particularly, for low a b s o r p t i o n laser m e d i a in which the efficiency is l i m i t e d by their small ab- sorption, just a double pass of the p u m p is enough to i m p r o v e the laser p e r f o r m a n c e [8]. Ideally, in- tracavity p u m p i n g would be an alternative m e t h o d to overcome the small a b s o r p t i o n problem, b u t due to the low p u m p i n g rate of the p u m p lasers it would n o t be feasible. By using a coupled cavities scheme, it is possible to increase the initial a b s o r p t i o n of the p u m p e d m e d i u m in order to increase the gain of the slave a n d still keep a low threshold for the cw op- e r a t i o n of the N d : YAG.

Supported by FAPESP under grant number 85/1629-8.

J On a CAPES-PICD fellowship. Present address at Universi- dade Federal da Bahia.

2. Theoretical

In order to analyse the b e h a v i o r o f the coupled

cavities scheme we assume that the spatial modes are matched. The main cavity consists of a 100% reflect- ing mirror (M1) and an output coupler ( M r ) o f r e - flectivity R, as shown in fig. 1. A second branch con- sists of an absorbing medium (with a transmission r) and a back mirror with reflectivity Ro. By sum- ming up all the electromagnetic contributions rein- jected into the main laser the equivalent reflectivity for the set R, Ro and r is given by [9]

Rely= [ ~ - - T r ~/R° exp (i~) ] 2 [ 1 - R ~ R o r e x p ( i ~ b ) ] 2 '

where ~ is the phase shift due to the optical length between mirrors R and Ro and due to the absorbing medium and the back mirror. For a constructive in- terference ( ~ = ( 2 m + 1 ) m m integer), the reflectiv- ity will be maximum, being in this case

Re.-= (2)

[ l + r ~ ] 2 "

Due to the increase in the refleetivity for the con- structive modes the p u m p laser will preferentially oscillate in the selected modes, but if the p u m p gain laser bandwidth is much greater than the cavity mode spacing there will be allways several modes that will

\

N o t : Y A G J L , ----.4 --.

\

Fig. I. Scheme of the coupling of the N d : Y A G laser to the TI ° ( 1 ) : KCI color center laser. The Nd : YAG laser cavity consists of mirrors M, and M2; mirrors M3, M4 and Mo are the CCL astig- matically compensated cavity. The spatial modes of both lasers overlap at M2. Prism P l , at Brewster angle, allows for the collin- earity of both beams at the KCI:TI°(1 ) crystal. The apertures inside the N d : Y A G laser are for TMoo operation. Mirrors M3, M4 and the crystal are in a v a c u u m chamber. M 3 and M 4 a r e gold coated mirrors. Prism P: intercepts the CCL beam, directing it to mirror Mo.

efficiently depopulate the gain medium.

Assuming now that the back mirror reflectivity is unity (Ro= 1 ) for the coupled mode, the main cavity threshold is lowered by a quantity Ag= ( 1 - R ) r /

( r + x / R ) , so lasing will occur before achieving the free resonator threshold gain go.

The transmission of the saturable absorber de- pends on the incident intensity according to

( l + l / I s ) - 1 , r = z o (3)

where Is is the saturation intensity.

At the begining of the operation, the main cavity gain is lowered by Ag(ro) until the laser action be- gins. At this point, the intensity grows such that />>Is; therefore r ~ l and the new threshold is go - ( 1 - R ) / ( 1 + ~ / R ) ,

and the system is now with its population inversion much higher than the cavity threshold population, initiating the Q-switched operation.

The photon life time in this coupled scheme de- pends entirely upon the ratio of energy transfer be- tween both cavities and on the saturable absorber transmission dependence on the time. The depen- dence of the total intracavity power, normalized by the m a x i m u m power, on the transmission of the sat- urable absorber is given by

P~P - ( 1 + x / R ) [1 + ( 1 2 + tfx/R)" ] , (4) where t = x / 7 . a n d 7"= 1 - R .

The time dependence of P~ is given by 1 OP~ ( 1 - x / R ) ( I + x / R ) 2 Or

- =

r - - .

P 0! ( 1 -k- r , ~ f R ) 3 Ol

At the m a x i m u m of the saturation, the population of the saturable absorber is fully inverted (for a 4- level scheme). The transmission at this point is given by r = 1 - a nl (a is absorption cross section, n is the ground state population and 1 is the length of the ab- sorbing medium ). The recovery of this population is exactly the decay of the total population no with an effective decay time td. Therefore, around r ~ 1 we have

0 r _ In To ( 6 )

Ot ta

Volume 79, number 1,2 OPTICS COMMUNICATIONS 1 October 1990 and consequently

1 0P~ 1 - ~ In ro

P 0t - 1 + ~ t~-- (7)

At the beginning of the intensity decay the behav- ior is exponential with a characteristic decay time tc given by

l+ k td (8)

tc= l - , ~ I n ro "

When R = I, the decay time is infinite (cw oper- ation) corresponding to an uncoupled cavity with=

out output coupling. At ro approaching unity (no ab- sorption), the cavities are coupled but, as there are no time dependent losses, we also obtain cw oper- ation. It is also easily seen that the first factor of the above expression takes into account the coupling be- tween the two cavities (the energy transfer term).

The second factor is solely dependent on the satur- able absorber physical parameters.

3. Experimental

KCI:T1 crystals were grown in our laboratory by the static Bridgman method using reagent grade starting materials sealed under argon in quartz am- poules. Single crystals 5 cm long with diameters of 2 cm were obtained with a T1 concentration of about 0.2 tool% measured by emission spectrography. This concentration is in the ideal range due to the fact that higher quantities of T1 impurities in the crystal gen- erates undesirable TI aggregates [ 10 ]. These samples underwent a thermal treatment of 650°C for 30 min- utes followed by a rapid quench to room tempera- ture to eliminate all the undersirable aggregation of T1 centers. This thermal treatment is rapid enough to avoid a considerable loss of substitutional TI ions through the sample's surface.

Samples of 10X 10X2 mm 3 cleaved and polished were irradiated on each face at 77 K by an electron beam accelerator with a current of 20 ~tA m i n / c m 2.

The obtained F center concentration was in the or- der of 1018 c m - 3 also within the ideal range to avoid F aggregates. After electron irradiation these samples were optically treated with light in the F center re- gion at - 30°C photoconverting F centers into TI°( 1 )

centers. Densities up to 10 j7 c m - 3 of the TI°( 1 ) were obtained under these conditions with an absorption band peaking at 1.04 ~tm, corresponding to an initial absorption of 90%.

Crystals of LiF were also obtained in our labora- tory by initially zone refining (under HF atmo- sphere) a reagent grade LiF starting material. The material thus purified was used to grow LiF crystals by Czochralski's pulling method under argon. Single crystal boules 30 mm of diameter and 80 mm of length were obtained and cleaved into plane samples of 1 0 X 1 0 X 2 mm 3. These samples were electron beam irradiated at room temperature during a time enough to produce a statistical concentration of F2 centers that caused an optical absorption of 20%

at 1.064 p,m.

The slave laser cavity is the usual three-mirror as- tigmatically compensated cavity where the pumped medium is placed in the tightly focusing region (Wo ~ 20 ~tm) and is designed to be able to operate at low temperatures. For the CCL operation at 77 K a special cryostat was constructed with a holding LN2 time of approximately 70 hours for a volume of 5 liters.

The pumping of the CCL with the N d : Y A G laser

was done using a collinear arrangment as shown in

fig. 1. In this figure it is shown the coupling of the

two laser cavities: the piano-concave N d : Y A G laser

back mirror M~ and the output flat mirror M2 are

aligned collinearly with the astigmatically compen-

sated CCL cavity (mirrors M3, M 4 and Mo). The

pump beam mode is matched exactly with the CCL

mode by coinciding the beam waists of both lasers

at mirror M2, since in the long arm of the CCL the

beam waist is almost exactly the waist of the

N d : Y A G laser (0.5 mm radius) at the output mirror

(see fig. 1 ). A prism at Brewster angle inserted in

front of the CCL vacuum chamber allowed for the

collinearity of both laser beams in the highly focused

region of the CCL cavity (between mirrors M3 and

M 4). This equilateral prism, made of dense flint glass

produced a dispersion of 1 m m / 1 0 cm for the 1.064

lain laser beam. Another equivalent prism was used

to intercept only the 1.5 lain beam deflecting it also

in the minimum deviation angle. The latter prism

directed the CCL beam into the CCL cavity output

mirror. It was used a CCL output mirror with a cur-

vature radius of 3 m thus allowing for the long arm

of the CCL to be stretched to an o p t i m u m experi- m e n t a l c o n d i t i o n ( ~ 80 cm ) a n d still keeping its sta- bility range in the 1 m m region ( a n d therefore m a t c h i n g the m o d e waist at the crystal). The o u t p u t m i r r o r t r a n s m i s s i o n of this m i r r o r (Mo) was 10% at

1.5 ~tm.

The coupling was tested using a d u m m y crystal in the focus of the CCL cavity. The laser threshold for the perfect coupling was essentially the operation threshold of the p u m p i n g lamp of the Nd: YAG laser, since this coupled cavity has no o u t p u t coupling b u t just m i n o r losses.

The M2 o u t p u t coupler reflectivity R of the N d : Y A G laser used was 88% for K C I : T I ° ( 1 ) a n d 72% for F y :LiF. In the latter case, m i r r o r M, has 12% t r a n s m i s s i o n at 1.064 ~am. F o r the F2 : LiF crys- tal a second a r r a n g e m e n t was made. As the emission b a n d overlaps the a b s o r p t i o n b a n d in the Nd laser emission line, t h e F y center firstly absorbs, and, after a delay time, emits in the same p u m p i n g wavelength;

therefore there is no need for a second cavity oper- ating at the peak of the F~ : LiF emission b a n d . We then simplified the a r r a n g e m e n t by e l i m i n a t i n g the prisms.

4. Results and discussion

T h e coupling scheme was tested for both centers described above a n d the results are

( A ) T I ° ( I ) : K C I center case

In this case, the N d : Y A G laser m i r r o r M~ is 100%

reflective, M2 has 12% t r a n s m i s s i o n at 1.064 lain. The m a i n cavity gain threshold is ~ 6.4% per pass a n d coupled with the CCL cavity the n o n saturated gain threshold is reduced to 5.2% per pass. At the satu- rated regime the threshold gain is just enough to c o m p e n s a t e the residual losses. In this case, the net gain m o d u l a t i o n for the n o n saturated a n d saturated regime is 5.2%, almost 3 times greater than in the F~- center case. We observed that in this case the me- chanical coupling is more critical a n d for p u m p pow- ers above the m a i n cavity threshold a m i s a l i g n m e n t leads to a m i x t u r e of pure cw a n d Q-switched pulses.

By a j u d i c i o u s fine a l i g n m e n t , pure Q-switched op- eration can be achieved, T h e l a m p o p e r a t i o n current at laser threshold was lowered from 24 A to 20 A.

The threshold of the CCL was the same as of the

1 , 5 E I i I E

g

0.0 ~ / ~ / L t =

85 27 219 31

Kw LANP CURRENT (A)

Fig. 2. The figure shows the power output of the TI°( 1 ) center, in the coupled configuration of fig. 1, as a function of the pump- ing Kr lamp of the Nd:YAG laser. The observed non linearity is due to thermal effects in the Nd:YAG rod. In the insertion it is shown the pulse width of the Nd:YAG leaking through the back mirror (T~0.2%) at 20 A.

N d : Y A G laser. The laser was chopped with a low duty cycle to avoid thermal effects. The typical pulses are 1 p,s long, as shown in the insert of fig. 2. The o u t p u t power of the laser as a f u n c t i o n of the lamp current is also shown in fig. 3. The n o n linearity ob- served is due to the thermal effect in the laser rod

,5

- 3

2"0 l . . . / ~ ' 2W

1.5

1

2 0 2 2 2 4

2 6 2 8 3 0 3 2

K r

Lomp

(A)

Fig. 3. Operation parameters of the Nd:YAG laser coupled with the CCL cavity of fig. 1 without the prisms, with the F2 :LiF crystal. The coupling mirror M2 has 28% transmission at 1.064 lain and the output coupling has 12% transmission (mirror M~ ).

The output power (square dots), the pulse frequency (circular dots) and the pulse width FWHM (crosses) are shown as a func- tion of the Kr lamp pump current.

20KHz

g

i . ~ l x

OKHz ~ D_ l.,a ('-

~g

g

Volume 79, number 1,2 OPTICS COMMUNICATIONS 1 October 1990 [12]. At the maximum lamp current used (29 A),

the time average laser output was 1.3 W, correspond- ing to an absolute efficiency of 40%. The output pulses of the CCL were also modulated as it could be seen by a slow detector (Ge detector frequency re- sponse ~ 100 kHz), but the actual pulse width was not measured.

Due to the fast stimulated decay rate in the color center laser, we expect that the CCL Q-switched pulses have approximately the same temporal pulse width ( 1 ~ts). Therefore, at the maximum pump lamp current (29 A), with a pulse rate of 25 kHz and an average power of 1.3 W, the energy per pulse is 52

~tJ and the output peak power is 52 W. The repeti- tion rate of the cw, Q-switched pulses was the same of the Nd laser as one would expect, and was 25 kHz at maximum lamp input power. The pulse frequency increases with the pump power.

In order to calculate the pulse width we have to consider that the total decay time is due to sponta- neous and stimulated emission and is given, for high intensities [13], by td=t~.o I/Is. The saturation in- tensity for the TI°( 1 ) center is 4 . 8 × 1 0 2 2 c m - 2 s - 1

and, considering the average power as 1.3 W, for an area of 1.2 × 10-5 cm-2, we obtain an average decay time of 94 ns. The calculated characteristic decay time tc in this case is 1.3 ~ts, very close to the 1 ~ts measured.

The remarkable fact in this operating scheme is that, as the color center is the unique effective source of loss for the Nd : YAG laser, the pumping efficiency is shown to be very high. The fact that the laser is operating in the Q-switched regime justifies this in- crease in efficiency since the population is inverted in a time that is a fraction of the decay time. Then, losses due to the excited state population fluores- cence are minimized.

order to increase the coupling factor, the transmis- sion of mirror M2 was increased to 28%. The main cavity threshold gain per pass in this case is 23% plus residual losses. By coupling the CCL cavity, the non saturated single pass gain threshold is reduced to 14.6% and the saturated one is reduced by an extra factor of 1.8% (this net gain modulation is very small). The uncoupled threshold gain is higher than the maximum possible laser gain so only the coupled laser cavity allows for laser action. The results are shown in fig. 2. The increase of the pump power re- duces the gain build up time; consequently, the pulse frequency increases proportionally, increasing the average output power. By increasing the pump power, the pulse width initially decreases, achieving a min- imum of ~ 1 ~ts. As the thermal lens effect in the N d : Y A G rod takes over, the cavity mismatch di- minishes the role of the saturable absorber and pro- duces longer pulses.

Considering that the F~- saturation intensity is 4 × 1023 c m - 2 s - I , and using expression (3), the ef- fective decay time is reduced to 35 ns, thus the char- acteristic photon decay time is 2 ~ts, much longer than the experimental value of ~ 1 ~ts. The reasons for this discrepancy are: firstly, the single pass gain variation during the saturation of the absorber is very small ( 1.8% ), what produces a slow build up of the pulses;

secondly the 12% coupling of mirror M~ presents an extra loss per pass that should account for the further reduction in the cavity photon life time [14].

The output power obtained is ~ 65% of nominal cw mode due to the high saturation intensity of the F~- color centers. These losses are due to the energy required to bleach the saturable absorber. Part of this power is lost by luminescence (the decay time is ten times faster than the pulse width).

(B) F ? : L i F center case 5. Conclusions

The F ~- centers are formed by electron beam ir- radiation in concentrations orders of magnitude smaller than other primary color centers, therefore the maximum center concentration is limited to small values. For a thickness of 1.74 ram, the absorption at 1.064 pm is 20%. Using mirror M2 with 12%

transmission, the coupling was very weak, producing long Q-switched pulses (several microseconds). In

It was shown that by coupling two cavities, a cw

Q-switching of the Nd: YAG laser was obtained due

to the injection mechanism of the beam after passing

through the color center medium. Secondly, the

unique loss mechanism of the Nd laser is the ab-

sorption of the TI°( 1 ):KC1 centers implying, there-

fore, a very efficient pumping scheme. In the

F~- :LiF case, the stimulated emission of the F~- cen-

t e r o c c u r s at t h e p u m p i n g w a v e l e n g t h d u e t o i t s s m a l l S t o k e s s h i f t . T h e r e f o r e , in t h i s c a s e , a s i m p l e c o u p l e d c a v i t i e s a r r a n g e m e n t w a s e n o u g h t o p r o d u c e Q - s w i t c h e d p u l s e s o f e s s e n t i a l l y t h e s a m e p u l s e w i d t h

( 1 g s ) a n d w i t h c o n s t a n t a m p l i t u d e .

References

[ 1 ] L.F. Mollenauer, in: Tunable lasers, eds. L.F. Mollenauer and J.C. White (Springer, Berlin, Heidelberg, 1987 ) p. 225.

[2] T.T. Basiev, S.B. Mirov and V.V. Osiko, IEEE J. Quantum.

Electron. 24 (1052) 1988.

[3] L.F. Mollenauer, N.D. Vieira and L. Szeto, Phys. Rev. B 27 (1983) 5332.

14] W. Gellerman, F. Luty and C.R. Pollock, Optics Comm. 34 (1981) 391;

L.F. Mollenauer, N.D. Vieira and L. Szeto, Optics Lett. 7 (1982) 414.

[5 ] W. Gellerman, A. Muller, D. Wandt, S. Wilk and F. Luty, J. Appl. Phys. 61 (1987) 1297.

[6] Y.L. Gusev, S.I. Marennikov and S.Y. Novozhilov, Sov. J.

Quantum Electron. 8 ( 1978 ) 960.

[7] T.T. Basiev, Yu K. Voroniko, S.B. Mirov, U.V. Osiko, A.N.

Prokhorov, M.S. Soskin and V.B. Taranenko, Soy. J.

Quantum Electron. 12 (1982) 1125.

[8] K.R. German, U. Diirr and W. Kunzel, Optics Lett. I 1 (1986) 12.

[ 9 ] S.M.J. Kelly, Optics Comm. 70 (1989) 495.

[ 10 ] F.J. Ahlers, F. Lohse, J.M. Spaeth and L.F. Mollenauen Phys.

Rev. B28 (1983) 1249.

[ 1 I ] H.W. Kogelnik, E.P. Ippen, A. Dienes and C.V. Shank, IEEE J. Quantum Electron. 8 (1972) 373.

[ 12 ] W. Koechner, Appl. Optics 9 (1980) 2548.

[ 13 ] W. Koechner, ed., in: Solid state laser engineering (Springer.

New York, 1976) p. 83.

[ 14] L.V. Tarasov, in: Laser physics (Mir Publishers, Moscow,

1983) p. 296.